Summary On 07 July 2003, approximately 0958 eastern daylight time, a Beech 58TC Baron aircraft (registrationN6058T, serial numberTK-110) crashed into Lake Ontario, Ontario, approximately 3nautical miles southeast of the Toronto City Centre Airport. The privately owned and operated aircraft was carrying out a LOC/DMEB instrument approach to Toronto City Centre Airport, after a flight from Lansing Municipal Airport, Chicago, Illinois. When the aircraft did not arrive at the airport and failed to respond to transmissions from the tower, a search was commenced. Patchy fog in the area resulted in ceilings variable from zero to unlimited and visibility from 1/8mile to more than a mile. Several hours later, the Metropolitan Toronto Police Marine Unit found debris on the surface of Lake Ontario. The aircraft was located the following day by the Ontario Provincial Police, using a sidescan sonar. The aircraft was essentially intact, resting vertically on its nose at a depth of 220feet. The deceased pilot was located in the aft cabin of the aircraft. He received minor injuries in the impact but failed to egress the aircraft for unknown reasons and died as a result of drowning. Recovery of the aircraft was technically difficult because of the depth of the wreckage. After some delays in the recovery operation, and under the auspices of the Metropolitan Toronto Police, the aircraft was raised on 21July2003. Ce rapport est galement disponible en franais. Other Factual Information History of the Flight The pilot was flying to Toronto to attend a business meeting scheduled for the afternoon of 07July2003, in downtown Toronto. He got a weather briefing at 2030 central daylight time (CDT)1 the day before the flight and filed an instrument flight rules (IFR) flight plan for a departure at 0700CDT the following morning. On 07July2003, the pilot got a weather briefing at 0512 CDT and departed Lansing at approximately 0645 CDT. The aircraft proceeded on a direct route toward Toronto at 13000feet, crossed into Canadian airspace in the vicinity of Sarnia at 0900 eastern daylight time (EDT), and received revised routing via the St.Catharines non-directional beacon to the Toronto City Centre Airport (TCCA). At 0917, in the vicinity of London, the pilot began a descent to 7000feet above sea level (asl) and was later cleared to 4000feet and to the TILEL fix2 to hold southeast on the localizer. N6058T reached the holding fix at 0948 and entered the hold, as shown in AppendixA. N6058T was given an expected further clearance time of 1400Z (1000EDT) because of an aircraft ahead flying the LOC/DMEB3 approach to Runway08. At 0950, the aircraft ahead reported on the missed approach. The flight crew indicated that they had visual contact with the airport but were unable to complete the approach because only part of the runway was in sight. The pilot of N6058T acknowledged hearing this report. Toronto Terminal advised him that the latest weather was visibility mile with a few clouds at 8800feet (indicated by an automated weather observation system [AWOS]). The pilot requested an approach to TCCA. Toronto Terminal cleared N6058Tfor a LOC/DMEB, circling for Runway08, to cross the TILEL fix at 3000feet asl. As N6058T passed the fix, the pilot was instructed to change to tower frequency. Tower cleared the aircraft to continue the approach and to report the runway in sight or in the missed approach. Three minutes later, the Tower advisedN6058T that runway visual range (RVR) was greater than6000, wind calm, and it then cleared the aircraft to land with the option of landing on Runway26, 33or08. The pilot acknowledged the landing clearance; this was the last transmission fromN6058T. After passing the VOKUB5DME fix, the aircraft continued inbound on the localizer and continued the descent below the minimum descent altitude (MDA) of 760feet asl until it struck the water, essentially on the localizer at approximately3.6DME. The aircraft was recovered from the lake and a preliminary inspection was conducted. The wreckage was transported to the Transportation Safety Board (TSB) Ontario Regional examination facility for further examination. Technical assistance was provided by Raytheon Aircraft Company, Teledyne Continental Motors and Hartzell Propeller Inc. Various instruments and indicators were sent to the TSB Engineering Branch Laboratory for a detailed examination. Radar and Communication Data Data was obtained from a radar located at Toronto/LesterB. Pearson International Airport (LBPIA), 14miles northwest of the TCCA. Track and altitude derived from the radar data is presented in AppendixA. The aircraft maintained an altitude of 4000feet asl and an airspeed of approximately 140knots4 while entering the hold. When cleared for the approach, the pilot began an immediate descent and airspeed increased to approximately 190knots. The aircraft crossed TILEL, descending through approximately 3500feet asl. At 13.5DME, the aircraft levelled at 2900feet and the airspeed decreased to 150knots. The aircraft resumed its descent at 11DME, initially with a rate of descent of approximately 1000feet per minute, then reducing to 600-700feet per minute with an airspeed reduction to approximately 120knots. At 6.5DME, the rate of descent began to increase again and the airspeed increased to approximately 150knots. According to radar, the aircraft passed the 5DME fix at exactly 1000feet. The last radar hit was at 500feet asl, with a rate of descent of approximately 1200feet per minute and an airspeed of 150knots. The geographic position of impact was consistent with the radar data. Communications records indicate that N6058T was in contact with Cleveland Centre until entering Canadian airspace near Sarnia at 0900. From that point on, the aircraft was in communication in succession with Toronto Centre, Toronto Terminal, and City Centre Tower. On a few occasions, transmissions from N6058T were clipped or broken. Otherwise, communications were normal and the pilot reported no difficulty of any kind. At 0825, while still in U.S. airspace, the pilot contacted Chicago Flight Watch and received a Toronto weather update. While in Canadian airspace, there was no record of N6058T contacting any flight service station within range for further weather updates. The 0900 (1300Z) weather was broadcast on the automatic terminal information service (ATIS) for the TCCA, but it cannot be determined if the pilot received it. Pilot Injuries and Survival Aspects The pilot suffered minor, non-lethal injuries in the impact and died by drowning. There was a contusion on the outside of the left hip, which is consistent with pressure against the lap belt during impact. There were no significant injuries consistent with the pilot striking the control column or instrument panel. Therefore, it was concluded that the pilot's lap belt and shoulder harness were secure at impact. The harness buckle was open, and the pilot was found free of the harness in the aft of the fuselage behind the left, front seat. There was no damage to the webbing or hardware on the seatbelt, and it was concluded that the pilot unlatched the buckle after the impact. Toxicological examination indicated nothing remarkable. The right front seat, including the seat rails, was no longer attached to the aircraft structure. The seat had failed from a force directed upward and aft against the seat pan, which indicated damage due to the ingress of water. The left front seat was still attached to the aircraft structure and in a normal position for flight. Nothing physical was found that would have impeded the pilot's egress through the passenger door, which had opened by itself during the impact. The pilot failed to egress for unknown reasons. Pilot Information The pilot held a private pilot licence and instrument rating issued by the U.S. Federal Aviation Administration (FAA). He was originally licensed in1960. He had owned and operated N6058T since1992, and had accumulated approximately 700hours of flight time in it. He had a medical certificate dated 16July2002, valid until 31July2003. Records indicate that he had a history of atherosclerotic heart disease with a myocardial infarction in January2000. His recovery was reviewed by an FAA medical specialist, and he was found medically fit for flight. The condition was not a factor in this occurrence. He took some prescription medications, none of which were an influence in this occurrence. The pilot had flown a biennial flight review and instrument rating proficiency check on 05July2003, two days prior to the accident flight. During the check, the pilot flew two instrument landing system (ILS) approaches and one global positioning system (GPS) circling approach, in visual meteorological conditions (VMC). Although the autopilot could have been engaged for the approaches, they were all flown manually. His handling of the airplane was normal, with the exception that he lowered the landing gear and flaps earlier than normal in order to slow the airplane without reducing power and risking shock cooling of the engines. Apart from some minor procedural errors and, on one occasion, poor attention to pitch attitude resulting in slow airspeed, the flight was assessed as being satisfactorily flown. Records indicate that the pilot flew 15hours in the previous 90days. During that time, he did not log any flight time in instrument meteorological conditions (IMC) or any actual or practice instrument approaches. Records back to the beginning of2002 indicate three encounters with actual instrument conditions, each involving a non-precision approach at Lansing, the pilot's home field. In addition, two instrument training flights were flown in one day, 10months prior to the accident. These training flights included four ILS approaches at locations near Lansing, and two non-precision approaches at Lansing. Aircraft Information The aircraft was a Beech58TC Baron with Teledyne Continental TSIO-520-WB engines and HartzellPHC-J3YF-2UL three-bladed propellers. It was equipped with a Garmin 530GPS, with moving map and integrated navigation and communication radios. The autopilot was a KFC200 flight director/autopilot; it did not have an altitude preselect feature. The aircraft was not equipped with a radio altimeter, a flight data recorder or cockpit voice recorder; neither recorder was required by regulations. Maintenance records for the aircraft, engines and propellers indicate that the aircraft was maintained in accordance with regulation, and there were no outstanding snags that would affect the airworthiness of the aircraft. The only apparent discrepancy with the aircraft was that the pilot's push-to-talk switch appeared to be intermittent, resulting in clipped transmissions. The Beech Baron58TC landing gear indicating system has three green lights that illuminate when the appropriate gear is down and locked. There is also a red light that illuminates any time that at least one gear is in transit or in an intermediate position. When all of the landing gear are up and locked, all of the indicating lights are extinguished. The intensity of the lights is automatically lowered when the navigation lights are turned on. The Pilot's Operating Handbook5 (POH) states that the landing gear position lights may not be visible in daylight when the navigation lights are on; it advises momentarily turning off the navigation lights in order to check the landing gear position. Damage to Aircraft Examination of the wreckage indicated that the aircraft struck the water in a wings-level, nose-level attitude. The landing gear was down and flaps were extended to 15degrees, the approach configuration. The three landing gear detached from the aircraft during the impact. The right-side door of the aircraft was forced open on impact and jammed against the right engine cowl, past the normal open position. Water was able to enter the aircraft through torn floor panels and the open door. The aircraft sank and came to rest vertically, on its nose, at a depth of 220feet. The fuselage, wings and empennage received very little damage. There were no indications of pre-impact structural anomalies that would have affected the controllability of the aircraft. All control surfaces were in place with little damage, and all control cables were continuous. Both engines were damaged as a result of the impact and/or water immersion. There was no indication of pre-impact faults or anomalies that would have prevented the engines from operating normally and producing their rated power. The blades on both propellers were bent aft and twisted toward low pitch, suggesting considerable rotational energy at impact. It was not possible to determine precise blade angle or power output at impact, however, the propellers were not feathered. The similarity of damage to both propellers indicated that both engines were developing approximately equal amounts of power at impact. The airspeed indicator dial face had a paint smear indicating 149to 154knots at impact. The vertical speed indicator had two smears, one indicating 1800feet per minute down and the other 2250feet per minute down. The altimeter sub-scale was set at 29.90inches of mercury; the altitude reading at impact could not be determined. Some abnormal switch, circuit breaker, and light positions and indications were noted. The aircraft battery switch was found on. The left alternator was on, but the left alternator circuit breaker had tripped, and the right alternator was off. Neither of the alternator warning lights showed evidence of being illuminated at impact. Both navigation lights were on at impact, indicating that electrical power was available. The ice light switch was on although conditions were daylight, icing conditions were not present, and all anti-icing switches were off. It is possible that these anomalies were the result of impact forces. The fuel boost pump switches were both off and the propeller synchrophaser switch was on; these positions are opposite to those called for in the POH before landing checklist. The right cowl switch was found in the closed position, and the left one was found in the open position; both were consistent with the actual position of the cowl flaps. Filament stretching indicated that both engine cowl lights were on at impact, suggesting that both cowl flaps were open. According to the POH, the cowl flaps are closed for descent and landing, and it was known that the pilot was habitually careful to avoid over-cooling the turbo-chargers. None of the landing gear position lights showed indications of being illuminated at impact; however, the landing gear were in the extended position when they were torn off at impact. The landing gear switch was found in the down position with the lever broken off in an upward direction due to impact, indicating that it had been in the down position prior to impact. None of the related landing gear circuit breakers were tripped. It is possible that impact forces were insufficient to cause filament stretch, with the landing gear indicating lights being illuminated at low intensity with the navigation lights on. Meteorological Information The weather was generally below forecast and was dominated by surface-based fog that was variable and patchy. The flight was planned the previous night based on an outlook of generally visual flight rules (VFR) conditions in the Toronto area, with the possibility of some thunderstorms. When the pilot checked the weather in the morning prior to his departure, a forecast had not yet been issued for TCCA. He was given the forecast for Toronto/LBPIA, which indicated generally VFR with a lowest condition of ceiling 2000feet and visibility 2miles in thundershowers and mist. He also received the recent actual weather, which was consistent with the forecast. When issued approximately 30minutes later, the weather forecast for the TCCA was similar, with a lowest condition of ceiling 2000feet and visibility 2miles in thundershowers and mist. Beginning at about 0600, the AWOS at TCCA reported occasional ceilings as low as zero, and visibility diminished progressively between 0600 and 0700 to just over 1mile. An amended forecast was issued at 0724, indicating a temporary reduction in visibility to 1mile in light rain showers and mist with scattered cloud at 200feet until 1000. N6058T received this forecast from the Chicago Flight Watch at 0825. He also received the 0800 weather for TCCA: visibility 1miles in mist, ceiling 2700overcast, temperature20C. By 0816, TCCA weather had deteriorated to ceiling zero and visibility1/8mile. At 0900, the weather remained ceiling zero, visibility1/8mile. This weather was being broadcast on the TCCA ATIS when N6058T arrived in the Toronto area, but it could not be determined if the pilot had obtained the ATIS information. The forecast remained unchanged until 0946, when a new forecast indicated ceiling zero and visibility1/8mile, improving between 1000and 1200to visibility 2miles in mist, ceiling 800feet broken. N6058T was unaware of this forecast. TCCA control tower personnel observed some improvement beginning at about 0930 when RVR was 2800feet and the top of the CN Tower was visible. At that time, the aircraft ahead of N6058T began its approach. From 3000feet in the holding pattern in the vicinity of the TILEL 15DME fix, the crew of the aircraft ahead had visual contact with the lake surface. After passing the TILEL fix, the aircraft passed through some thin cloud layers and remained above a lower surface-based layer of ground fog. A spit of land at 3DME, to the right of the approach course, was not visible, but a lighthouse protruded through the fog layer, indicating a fog depth in the order of 50feet. At 2.2DME, the crew could see buildings and trees protruding through the fog on Toronto Island and some features of the City of Toronto. They had adequate visual reference to be able to carry out a circling procedure from 2DME, and cloud layers presented a good horizon. As the crew of the aircraft ahead made their turn onto final approach, the precision approach path indicator was briefly visible and the mid-portion of the runway was visible, but both ends of the runway were obscured by fog, resulting in the missed approach. At that time, conditions appeared to be almost VFR from the perspective of tower personnel, and RVR was6000. The AWOS issued a special report at 0944, indicating visibility mile, ceiling unlimited with a few clouds at8800feet. N6058T was given the 0944 weather report by Toronto Terminal before being cleared for the approach. About two minutes before the crash, while N6058T was conducting the approach, the control tower advised the pilot that the RVR was greater than 6000feet and gave him a landing clearance. One minute before impact, AWOS recorded visibility statute miles, ceiling 300feet broken, 8800feet overcast, temperature 19C, dew point 19C, altimeter29.90. Two minutes later, one minute after impact, an AWOS special report indicated wind calm, visibility 1statute mile, ceiling 200feet overcast, 8800feet overcast. Members of the Metropolitan Toronto Police Marine Unit reported that when they responded visibility was near zero, obscured by fog on the water surface, in the harbour and on Lake Ontario in the vicinity of the point of impact. Other nearby aerodromes (Toronto/LBPIA and Toronto/Buttonville Municipal) were also operational and had suitable weather for diversion had the pilot of N6058T requested it. N6058T had over 2hours worth of fuel remaining on the approach; this was adequate for a safe diversion in the event of a missed approach. Aerodrome and Navigation Aids The LOC/DMEB approach to TCCA uses the XTC localizer, frequency110.15, and the ITZ DME, channel38. The navaids were checked after the occurrence and found to be operating normally and within tolerances. A Notice to Airmen stated that the ITZ channel38 DME was unusable within 1.0DME. This affected the LOC/DME RWY08 approach but had no effect on the LOC/DMEB approach. The flight paths flown by the preceding aircraft and N6058T indicate that both were responding to accurate positioning. The TCCA also has the ILS/DME6 RWY08 precision approach to Runway08. An airspace conflict precludes the use of this approach when the Runway 23/Runway24 combination is in use at Toronto/LBPIA. Therefore, ILS/DMERWY08 approaches are authorized only when Toronto/LBPIA is using the Runway 05/Runway 06 combination. As a result, the ILS precision approach was not offered to N6058T. N6058T accepted the LOC/DMEB approach without requesting or otherwise indicating a preference for any other approach. Controlled Flight into Terrain Controlled flight into terrain (CFIT) is an occurrence in which an aircraft, under the control of the crew, is flown into terrain, water or an obstacle with no prior awareness on the part of the crew of the impending disaster.7 This type of accident can occur during most phases of flight, but CFIT is more common during the approach-and-landing phase. The Flight Safety Foundation8 noted that CFIT was the leading category of approach-and-landing accidents in a study of airline accidents and the second leading factor in all fatal business jet accidents. Its findings included the following: omission of action/inappropriate action by a flight crew member was the most common primary causal factor, usually referring to the crew continuing descent below MDA without adequate visual reference, either intentionally or unintentionally; lack of positional awareness in the air was the second most common factor, generally resulting in CFIT; 75percent of CFIT occurrences were associated with non-precision approaches, primarily when a precision approach aid was not available or was not used; significant terrain features were not necessarily a prerequisite for CFIT; a majority of CFIT occurrences were during poor visibility conditions; disorientation or visual illusions were involved in 21percent of occurrences, with a lack of vigilance, inadequate monitoring of primary instruments, and a lack of training and awareness identified as associated factors. An FAA study9 into CFIT accidents in general aviation had the following results: CFIT comprised almost one-third of general aviation accidents that occur in instrument conditions and cause 17percent of all general aviation fatalities; IFR-rated, general aviation pilots age 50and over have significantly more CFIT-type accidents than IFR-rated pilots under age50; differential sensory and cognitive capabilities, low annual flying hours, and inadvertent flight from VMC into IMC seem to contribute disproportionately to CFIT accidents. The FAA issued an Advisory Circular10 to promote awareness of CFIT in general aviation operations, citing the following awareness factors for general aviation operations involving IFR flight into IMC conditions: increased risk of CFIT with non-precision approaches; the importance of situational awareness; risk involved when transitioning from VMC to IMC or vice-versa; the importance of flying a stabilized approach; and additional risks when flying outside the U.S., including language, different depictions of terrain, elevation, and runway data. Instrument Approach Plate Design The pilot normally used FAA National Aeronautical Charting Office (NACO) Terminal Procedures Publications for instrument approaches in the United States. For the flight to Canada, he used the Canada Air Pilot, Volume4. Canada Air Pilot and the FAA/NACO Terminal Procedures Publications are generally similar in appearance; however, there are minor differences, including the presentation of altitudes. The LOC/DMEB chart for TCCA shows an altitude of 1000feet at the VOKUB5DME fix. In Canadian approach plates, an altitude displayed without a horizontal line above or below the numerical value means a minimum altitude. In FAA/NACO approach plates, this format means a recommended altitude; a minimum altitude would have a solid horizontal line beneath the numerical value. Approach Ban Canadian Aviation Regulations11 prohibit pilots from completing an instrument approach (certain circumstances excepted) beyond the final approach fix for a runway served by one RVR, if the RVR value is below1200. The RVR at TCCA was greater than 6000when N6058T conducted the approach.